Camouflaged Trade Agreements Elie Appelbaum York University Canada

Camouflaged Trade Agreements Elie Appelbaum York University, Canada Mark Melatos The University of Sydney, Australia Midwest Meetings 2012 St Louis, MO

The purpose of this paper • Apparent paradox: – – • Theoretical primacy of CUs vs Observed popularity of FTAs A new explanation: – – There is no paradox … Some observed FTAs may, in fact, be CUs in disguise! 2

CUs reconsidered • What constitutes an “optimal” CU? – Standard vs Generalised CUs. – SCUs are typically sub-optimal. • So, why are CETs observed in practice? – The WTO requires them. • Yes, but WHY? ! – GCU higher TW external tariff rates. 3

So what? • GCUs not explicitly ruled out by WTO law. • Article XXIV GCUs and FTAs are indistinguishable from one another. • So, maybe, some FTAs are, in fact, GCUs! • And, hence, CUs may play a bigger role in world trade than is currently believed. 4

Another question …! • Since GCUs are welfare superior (for members), why do we observe any SCUs? – Observability of members’ TA rankings – Verifiability of presence of CET. – SCU>FTA members must levy a CET – SCU<FTA members implement a GCU • Alternative explanation: costly contracting in TAs (Horn et al. , 2010). 5

And yet another … • Why are GCUs not prohibited by the WTO? – GCUs and FTAs are legally indistinguishable. – Thus, WTO must either prohibit both FTAs and GCUs, or neither. – But, FTAs contribute significantly to welfare. – WTO regulates CU external tariffs to indirectly regulate how they are chosen. • Article XXIV irony: FTAs undermine the WTO’s ability to limit CU monopoly power. 6

Contributions • Main contributions: – CUs disguised as FTAs. – CUs more significant than currently believed. • Additional contributions: – Endogenise the decision to levy a CET. – Theoretical justification for Article XXIV. – Cautionary note on empirical testing – FTAs provide a convenient ‘fig leaf’ to hide the true intentions of CU members. 7

The Model • • Imperfect competition & partial equilibrium. Markets are segmented; Cournot competition. 3 countries: 1, 2 “passive”, 3 “active”. TA types: {FTA, SCU, GCU}; transfers permitted between member countries; MFN tariffs. • 3 -stage game: 1. Countries 1 and 2 choose a TA type. 2. All countries choose optimal tariffs 3. All countries choose outputs in each market 8

Information assumptions • All elements of the game are common knowledge. • But, TA member welfare rankings are not verifiable by the WTO. • Tariffs: – The WTO can verify whether or not members levy a CET. – But WTO cannot verify that the observed tariff rates correspond to the observed TAs. 9

GCUs welfare dominate Proposition 1: W(GCU) > W(SCU) except when members symmetric … W(GCU) = W(SCU). Proposition 2: W(GCU) > W(FTA) always. • GCU members face a larger feasible tariff set. Simplistic, but still holds here … • Country 3’s tariff choice independent of the TA choice made by countries 1 and 2. 10

When are GCUs observed? Proposition 3: (linear D, identical costs) i. W(GCU) > W(FTA) > W(SCU) if member Ds are sufficiently asymmetric. ii. W(GCU) > W(SCU) > W(FTA) otherwise. • In (i), may be able to implement a GCU in defiance of Article XXIV. • In (ii), may implement SCU rather than GCU to avoid possible WTO sanction. 11

C=1/2 12

The WTO’s CET requirement (assume: linear D, identical costs, a 1>a 2) Proposition 4: Proposition 5: If a 1, a 2 sufficiently similar: else, 13

World welfare (assume: linear D, identical costs, a 1>a 2) Proposition 6: If a 1, a 2 sufficiently similar: WW(FTA) > WW(SCU) > WW(GCU) else, WW(SCU) > WW(FTA) > WW(GCU) 14

Conclusion • Conjecture: many FTAs are, in fact, GCUs in disguise. – GCUs are optimal for member countries. • WTO should (and does, via Article XXIV): – Discourage standing alone. – Encourage SCU and FTA. • The ability to form a FTA undermines the welfare-enhancing role of the CET – And makes GCU possible. 15